The cortisol awakening response (CAR) in toddlers: Nap-dependent effects on the diurnal secretory pattern

Cortisol levels in adults show a sharp decrease from mid-morning to mid-afternoon. Most toddlers take afternoon naps, which is associated with a less mature diurnal pattern characterized by a midday plateau in cortisol secretion. Napping in preschoolers produces a robust cortisol awakening response (CAR), which may account for such maturational differences. This experimental study extends prior work by examining whether the presence and timing of the nap-dependent CAR influences the diurnal cortisol pattern in toddlers.

Toddlers (n = 28; 13 females; 30-36 months) followed a strict biphasic sleep schedule (≥ 12.5h time in bed; ≥ 90 min nap) for ≥ 3 days before each of four randomly ordered, in-home cortisol assessments. For each assessment, saliva samples were obtained at morning awakening, ∼ 09:30, pre-nap, 0, 15, 30, 45, 90, and 135 min post-nap awakening (verified with actigraphy), and ∼ 19:30. On one day, children napped at their scheduled time, and parents collected saliva samples. On another day, children missed their nap, and parents collected saliva samples at matched times. On two other days, children napped 4h (morning) and 7h (afternoon) after awakening in the morning, during which time researchers collected pre- and post-nap saliva samples. Saliva was assayed for cortisol (μg/dl).

Three-level multilevel models were used to estimate the CAR and diurnal cortisol patterns in all four conditions. Compared to the no-nap condition (no observed CAR; b = -0.78, p = 0.65), we found a pronounced cortisol rise following the morning nap (b = 11.00, p < 0.001) and both afternoon naps whether samples were collected by parents (b = 5.19, p < 0.01) or experimenters (b = 4.97, p < 0.01). Napping in the morning resulted in the most robust post-nap cortisol rise (b = 10.21, p < 0.001). Diurnal patterns were analyzed using piecewise growth modeling that estimated linear coefficients for five separate periods throughout the day (corresponding to morning decline, noon decline, post-nap rise, post-nap decline, and evening decline). We observed a significant post-nap rise in cortisol values on the parent-collected afternoon nap (b = 3.41, p < 0.01) and the experimenter-collected morning nap (b = 7.50, p < 0.01) days as compared to the no-nap day (b = -0.17, p = 0.82). No other differences in diurnal profiles were observed between the parent-collected nap and no-nap conditions; however, toddlers had a steeper evening decline on the day of the morning nap compared to the parent-collected afternoon nap (b = 0.30, p < 0.05) and no-nap conditions (b = 0.27, p < 0.05).

These well-controlled findings suggest that the presence and timing of daytime naps influence the pattern of diurnal cortisol secretion in toddlers. They also provide support for the hypothesis that napping is the primary state driving the immature midday plateau in cortisol secretion, which becomes more adult-like across childhood. Prior studies of the diurnal cortisol pattern have employed a cubic model, and therefore, have not detected all possible variations due to napping. Our experimental data have important methodological implications for researchers examining associations between the slope of the diurnal cortisol pattern and developmental outcomes, as well as those utilizing afternoon cortisol reactivity protocols in napping children.